[GUIDE] Scripted Yosemite/Mavericks Install on Gigabyte Mobos

[damiafix]

Hi, Many thanks! at this point, also if OSX is very stable (no crash, very fast with the CPU at 3Ghz) I'll try with another installation following the instructions of this 3d. I was thinking that I could to install the gigabyte's kexts also from this installation.

 

Have a good day, D.

[digital_dreamer]

<snip>

I've upped again the CPU whit "easytune" from Window, until to 3.00ghz (the yellow step) but I've reduct the memory frequency to 1500mhz, and from the geekbench I've reached 10116, It's a good score for my setup?

The OSX now seem stable, but I wish to improve the performances, and to resolve the right recognition of the memory frequency and mine CPU model and frequency.

<snip>

That's a good score for a 3Ghz clock.

 

The CPU clock and model reporting in OS X is just cosmetic. So, no need to worry that the system is not running correctly. It'll run just as you have it set in the BIOS. The reporting in OS X is typically set via keys/strings in the smbios.plist. If you use the Chameleon RC5 bootloader (unofficial release), however, it can set many of these automatically. However, the keys in the smbios.plist will override them. Those keys/strings would need to be deleted (like SMmemspeed and SMcputype) for the bootloader's memory and CPU detection to be recognized in System Profiler.

 

Hope that helps.

 

Hi I have an ex58 extreme, I think it´s similar to ud5, I have installed the retail 10.6.3 with the guide of this post. But when I try to install the update to 10.6.4 the system crash when boot. I have tried with the script again post installation, editing the com.apple.boot.plist, with the flags -x -v cpus=16. I don´t know what to do. Please any ideas.

<snip>

As Sidewinder33 already mentioned, you likely need to use the pmVersion=20 kernel flag. You can enter this at boot time to get in. However, this needs to be added/updated in the boot plist. It may already be set to pmVersion=19 to work with 10.6.3

 

Sidewinder33,

About the RAID support files copying. Ouch! That's not good.

Let me look at it from my end and see what I can find. I may need your Install.log from the script, if I can't find anything here.

I haven't used the AppleRAID in a few months, so need to try again. I bounce back and forth from AppleRAID setups to non-RAID setups due to limited number of HDs on hand.

 

regards,

MAJ

[jatar]

That's a good score for a 3Ghz clock.

 

The CPU clock and model reporting in OS X is just cosmetic. So, no need to worry that the system is not running correctly. It'll run just as you have it set in the BIOS. The reporting in OS X is typically set via keys/strings in the smbios.plist. If you use the Chameleon RC5 bootloader (unofficial release), however, it can set many of these automatically. However, the keys in the smbios.plist will override them. Those keys/strings would need to be deleted (like SMmemspeed and SMcputype) for the bootloader's memory and CPU detection to be recognized in System Profiler.

 

Hope that helps.

 

 

As Sidewinder33 already mentioned, you likely need to use the pmVersion=20 kernel flag. You can enter this at boot time to get in. However, this needs to be added/updated in the boot plist. It may already be set to pmVersion=19 to work with 10.6.3

 

Sidewinder33,

About the RAID support files copying. Ouch! That's not good.

Let me look at it from my end and see what I can find. I may need your Install.log from the script, if I can't find anything here.

I haven't used the AppleRAID in a few months, so need to try again. I bounce back and forth from AppleRAID setups to non-RAID setups due to limited number of HDs on hand.

 

regards,

MAJ

 

I will try it tonight. What´s the value 20 or 19, Is ir refered to the CPU?

Thanks

[Thorgal]

Update to 10.6.4. kills my system - graphics related ?

 

Before you state the obvious : yes, I removed sleepenabler.kext and I can boot fine into 10.6.4, but the machine will hang after a couple of minutes.

 

Really don't have an explanation for this : everything runs perfect on 10.6.0 (fresh install), injected my GTX 280 into boot plist, working great with QE/CI enabled.

 

Fast forward to 10.6.4 update : graphics performance is literally cut in half with the update (Cinebench 11 : 13 fps instead of almost 26, Core Image test slowing to a crawl) , and as mentioned system does not stay up longer than about 5 minutes.

 

Rest of setup : Gigabyte EX58-Extreme + 6Gb memory + W3540, all running stock speeds.

 

Any ideas ?

[digital_dreamer]

I will try it tonight. What´s the value 20 or 19, Is ir refered to the CPU?

Thanks

20 for 10.6.4

 

MAJ

 

Now DD,

 

I've also come across another odd occurrence. When I run the kext installer (I believe option 5), it proceeds through installs the kext and updates the extensions.mkext. No trouble there, the issue I've been having is that a lot of the times its only updating one of the helper partitions in my RAID setup. I have to manually view the partitions and move the respected kexts and extensions.mkext over to the un updated partition. The thing that strikes me as odd is while its installing and updating the extenetions.mkext it says its installing it in both partitions. I noticed this after messing with some LAN kexts, I uninstalled one put another in and when I checked the partitions, one still contained the old one and the other the new one.

 

Jeff

Sidewinder33,

I just did a install on a 2 drive AppleRAID setup (stripe) and didn't encounter any issues. All files are updated on both helper partitions when I do a kext install or update boot cache, as expected. I installed different kexts and uninstalled them without it skipping a beat. This was for an /Extra install, not EFI.

I'll need your Install.log from the script to see what's going on.

 

best of wishes,

MAJ

[damiafix]

That's a good score for a 3Ghz clock.

 

The CPU clock and model reporting in OS X is just cosmetic. So, no need to worry that the system is not running correctly. It'll run just as you have it set in the BIOS. The reporting in OS X is typically set via keys/strings in the smbios.plist. If you use the Chameleon RC5 bootloader (unofficial release), however, it can set many of these automatically. However, the keys in the smbios.plist will override them. Those keys/strings would need to be deleted (like SMmemspeed and SMcputype) for the bootloader's memory and CPU detection to be recognized in System Profiler.

 

Hope that helps.

 

Many thanks, I'll try tomorrow and I let you know. Do you think that is really necessary to install again OSX? Can I use your scripts after a Iaktos installation?

 

Best regards, D.

[jatar]

I erased the sleepenabler kext from the extra folder. I modified the com.apple.boot.plist with pmVersion=20 and again kernel panic.

 

 

My com.apple.boot.plist:

 

<?xml version="1.0" encoding="UTF-8"?>

<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">

<plist version="1.0">

<dict>

<key>Kernel</key>

<string>mach_kernel</string>

<key>Timeout</key>

<string>3</string>

<key>Kernel Flags</key>

<string>arch=x86_64</string>

<string>cpus=16</string>

<string>pmVersion=20</string>

<key>device-properties</key>

<string>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<string> <key>EthernetBuiltIn</key>

<string>y</string>

</dict>

</plist>

 

 

Is ti ok?

Thanks again.

[digital_dreamer]

Many thanks, I'll try tomorrow and I let you know. Do you think that is really necessary to install again OSX? Can I use your scripts after a Iaktos installation?

 

Best regards, D.

I don't think a reinstall is really necessary. However, we don't know what kind of install the Laktos S2 V3 did. What modified kexts were installed into S/L/E? What kexts are installed in /Extra? What kind of install is it: does it have a "/Extra/Extensions" folder or is there a boot cache (mkext file)?

A list of the /Extra directory (and S/L/E) contents will help us see how it's being used and what may be done to make it work with the script.

 

Basically, the script can attempt a install, but the existing setup may create a conflict. For example, if the existing install uses a /Extra/Extensions folder, that will need to be deleted, as the script requires a boot cache setup.

 

You could just rename your /Extra folder to "Extra-original" and let the script create a new setup for you.

 

MAJ

 

<string>cpus=16</string>

<string>pmVersion=20</string>

jatar,

(Would you edit your post and snip that device-properties line so that the code doesn't overflow horizontally?)

 

Delete the above quoted strings, as, not only are they not needed, they are not properly set up or placed.

All Kernel Flags arguments are to be in the same string, separated by a space. It would be like the following:

<key>Kernel Flags</key>

<string>arch=x86_64 cpus=1 pmVersion=20</string>

 

The cpus= argument is for specifying the number of cores/CPUs the kernel will use. It is typically used to limit the cores used to just one, if the DSDT file has processor aliases that need to be edited out, like on many ASUS boards. Your argument specifies more than what's available. I don't know what the implications are of that, but it can't be good.

 

The pmVersion= argument is no longer needed if you removed the sleepenabler.kext.

 

best of wishes,

MAJ

[Thorgal]

Update to 10.6.4. kills my system - graphics related ?

 

Before you state the obvious : yes, I removed sleepenabler.kext and I can boot fine into 10.6.4, but the machine will hang after a couple of minutes.

 

Really don't have an explanation for this : everything runs perfect on 10.6.0 (fresh install), injected my GTX 280 into boot plist, working great with QE/CI enabled.

 

Fast forward to 10.6.4 update : graphics performance is literally cut in half with the update (Cinebench 11 : 13 fps instead of almost 26, Core Image test slowing to a crawl) , and as mentioned system does not stay up longer than about 5 minutes.

 

Rest of setup : Gigabyte EX58-Extreme + 6Gb memory + W3540, all running stock speeds.

 

Any ideas ?

 

Tried all night to get it stable, no luck... Only other periferal attached are a wireless Logitech mouse, anyone knows of reports on troubles with that after update. Seems to be graphics related though : if I recall correctly Apple updated graphics drivers going from 10.6.3 to 10.6.4 ? Not using any other kexts than standard ones (used Aquamac's method to inject, according to him nothing else is needed, and this is true at least for 10.6.0).

 

I guess it's back to original install for me, unless anyone here has something extra I can try (or point me towards my own stupidities )

 

PS: DD, the latest script also does not allow me to rebuild kext caches when installing, I have to do this manually through single user mode, (and be quick before the system panics).

[jatar]

Well I retouched ths boot.plist as digital dreamers says me and all it´s ok.

All problems are to have not the flags in the same line?

 

boot.plist:

<?xml version="1.0" encoding="UTF-8"?>

<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">

<plist version="1.0">

<dict>

<key>Kernel</key>

<string>mach_kernel</string>

<key>Timeout</key>

<string>3</string>

<key>Kernel Flags</key>

<string>arch=x86_64 pmVersion=20</string>

<key>device-properties</key>

<string>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<string> <key>EthernetBuiltIn</key>

<string>y</string>

</dict>

</plist>

 

 

 

 

Well, thanks a lot of, I have now minor issues with reboot and shut down, but they are not important.

Saludos

[Sidewinder33]

Well I retouched ths boot.plist as digital dreamers says me and all it´s ok.

All problems are to have not the flags in the same line?

 

boot.plist:

<?xml version="1.0" encoding="UTF-8"?>

<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">

<plist version="1.0">

<dict>

<key>Kernel</key>

<string>mach_kernel</string>

<key>Timeout</key>

<string>3</string>

<key>Kernel Flags</key>

<string>arch=x86_64 pmVersion=20</string>

<key>device-properties</key>

<string>

<key>EthernetBuiltIn</key>

<string>y</string>

</dict>

</plist>

 

Well, thanks a lot of, I have now minor issues with reboot and shut down, but they are not important.

Saludos

 

Good to here Jatar, there are a few different things you can do to fix restart and shutdown, with kexts, or DSDT editing.

 

Thorgal

 

10.6.4 had a huge impact on graphics performance, there has since been a graphics update released from apple that seems to have restored graphics performance. Also maybe try using verbose boot, -v at the chameleon boot prompt. It should say what is causing the panic.

[digital_dreamer]

Tried all night to get it stable, no luck... Only other periferal attached are a wireless Logitech mouse, anyone knows of reports on troubles with that after update. Seems to be graphics related though : if I recall correctly Apple updated graphics drivers going from 10.6.3 to 10.6.4 ? Not using any other kexts than standard ones (used Aquamac's method to inject, according to him nothing else is needed, and this is true at least for 10.6.0).

 

I guess it's back to original install for me, unless anyone here has something extra I can try (or point me towards my own stupidities )

 

PS: DD, the latest script also does not allow me to rebuild kext caches when installing, I have to do this manually through single user mode, (and be quick before the system panics).

Did you not try the Graphics Update for 10.6.4 released by Apple? This was supposed to restore the many performance-related issues on nVidia cards.

 

As for boot caches not rebuilding, can you provide me more details here? Perhaps the Install.log would help. It's hard to imagine the "kextbuild" script working better than the main one, unless you are working from a Leopard install or trying to install a kext into /Extra that isn't specified to contain any boot cache data to begin with, like the many Networking kexts.

 

MAJ

[Thorgal]

Did you not try the Graphics Update for 10.6.4 released by Apple? This was supposed to restore the many performance-related issues on nVidia cards.

 

As for boot caches not rebuilding, can you provide me more details here? Perhaps the Install.log would help. It's hard to imagine the "kextbuild" script working better than the main one, unless you are working from a Leopard install or trying to install a kext into /Extra that isn't specified to contain any boot cache data to begin with, like the many Networking kexts.

 

MAJ

 

Call me stupid, but no I didn't try the graphics update Humming along on 10.6.3 right now and everything working fine, I see your magic still works as good as it used to MAJ, thank you !

 

Got some more info for you on the kext cache build. It's only a problem when installing to the EFI partition, which I still did because of old times sake I guess, don't know. Have now tried with the /Extra install and no trouble at all: cache build works perfectly through the script.

 

The errors I get when going for /EFI is that kexts are not authenticated, so cache build fails. When booting through -s and command buildcache all is fine though, if you're quick enough to issue the command before KP.

 

Anyway, not a big issue as you can see, I'd advise just to go with the /Extra install.

[Sidewinder33]

Hey DD Do you want my whole install.log or any one specific part? I'll happily attach either

[digital_dreamer]

The errors I get when going for /EFI is that kexts are not authenticated, so cache build fails. When booting through -s and command buildcache all is fine though, if you're quick enough to issue the command before KP.

Okay, thanks for that detail. Would you have a Install.log with that info included so I can look into it? It may still be part of your existing log.

 

Hey DD Do you want my whole install.log or any one specific part? I'll happily attach either

Whole log, zipped, would be okay.

Thanks!

 

MAJ

[Sidewinder33]

Okay, thanks for that detail. Would you have a Install.log with that info included so I can look into it? It may still be part of your existing log.

 

 

Whole log, zipped, would be okay.

Thanks!

 

MAJ

 

Here ya go DD, I pulled it using the script, Install.log.zip

 

Another thing I've noticed is that when I install a kext (using option 5) it resets the second of my two helper partitions. It deletes my dsdt and resets my themes to the default, instead of just copying from the other helper partition.

 

Also I know you've addressed this issue earlier but what lines would I need to remove to skip the duplicate mount point warning? Any time I don't select my raid as the install target I get the multiple mount point warning.

[Thorgal]

Okay, thanks for that detail. Would you have a Install.log with that info included so I can look into it? It may still be part of your existing log.

 

Sorry MAJ, I wiped the drive before reinstalling with /Extra method, as I wasn't sure what the problem was when I got stuck with 10.6.4, and started from scratch again. Next install I'll provide the log. Sorry again.

 

Edit scratch that, seems I copied the hackinstaller folder with the original log in it. It's a bit abridged (since I reinstalled 3 or 4 times), but I think the section that interests you is there.

 

Kind regards,

 

Thorgal

Install_log.zip

[digital_dreamer]

Here ya go DD, I pulled it using the script, Install.log.zip

 

Another thing I've noticed is that when I install a kext (using option 5) it resets the second of my two helper partitions. It deletes my dsdt and resets my themes to the default, instead of just copying from the other helper partition.

 

Also I know you've addressed this issue earlier but what lines would I need to remove to skip the duplicate mount point warning? Any time I don't select my raid as the install target I get the multiple mount point warning.

Okay, thanks for that log!

 

Now, the reason for the RAID partition issues appear to all point to that duplicate mount point created from "Macintosh " (There are two spaces on the end of that name). This is why the RAID files are not copying over to the second helper partition, as the duplicate mount point tricks the system into thinking it's already mounted. Therefore, all files get copied to that false mount point, instead of the real "Macintosh " RAID helper. This is also why the DSDT files and themes get "deleted" when doing a kext install.

 

So (and this is why I have that duplicate mount point warning in place), we need to delete that mount point. Here's what to do:

1. Start script - this will run a initiation routine to unmount all RAID helper partitions. Nothing else needs to be done in the script - just start it.

Just to verify for peace of mind, you do mount in Terminal and see if that partition, "Macintosh ", is still mounted. It shouldn't be listed. However, if you do ls /Volumes, it will show up. This shouldn't happen and only happens because it's a duplicate. This is how the script determines what is a duplicate or false mount point.

2. Open another Terminal window and do sudo rm -rf "/Volumes/Macintosh " (I'm putting this path in quotes to include the 2 spaces)

This will delete the false mount point.

3. Quit and relaunch the script. Things should work as expected, now.

 

On a side note, I notice you have some undeleted trash on one of the RAID helpers, "Macintosh 1"

The log shows: (Trash size: 72.7M)

This isn't a problem now, but eventually, if you continue to delete files without emptying the trash, you'll run out of space on that partition and RAID copies will fail, prompting one to wonder what happened. This happened to me many times. LOL! That's why the script keeps track of the trash and partition sizes. You'll have to View that helper Partition for the trash to show the contents and allow you to empty it. At any rate, the script will inform you if the copies failed and revert to original backup files.

 

In the script Preferences, one can disable the "Always check for duplicate mount points" option. I strong suggest not doing so, unless one strongly feels the false mount point is, err, false. This is the only real warning the script provides that indicates a serious issue affecting file access. All sorts of things break with this issue not being resolved.

 

Now, I'm going to look into why the Duplicate Mount Point warning didn't appear when you had the RAID volume selected as the target. I'll get this fixed so that it'll be even more annoying.

Unfortunately, there really isn't any other way around this mount point issue. When you're mounting/unmounting volumes many times and something goes wrong (crash, unexpected quit in the middle of an operation, etc.), the mount point doesn't get cleared properly and gets stuck there, even after a reboot. It has to be cleared manually. And, there is no way I will implement a routine in the script, as simple as it would be, to automatically delete a false mount point (like rm -rf /Volumes/Hard_Drive). That's asking for too much trouble. Ha!

 

best of wishes!

MAJ

[Sidewinder33]

Okay, thanks for that log!

 

Now, the reason for the RAID partition issues appear to all point to that duplicate mount point created from "Macintosh " (There are two spaces on the end of that name). This is why the RAID files are not copying over to the second helper partition, as the duplicate mount point tricks the system into thinking it's already mounted. Therefore, all files get copied to that false mount point, instead of the real "Macintosh " RAID helper. This is also why the DSDT files and themes get "deleted" when doing a kext install.

 

So (and this is why I have that duplicate mount point warning in place), we need to delete that mount point. Here's what to do:

1. Start script - this will run a initiation routine to unmount all RAID helper partitions. Nothing else needs to be done in the script - just start it.

Just to verify for peace of mind, you do mount in Terminal and see if that partition, "Macintosh ", is still mounted. It shouldn't be listed. However, if you do ls /Volumes, it will show up. This shouldn't happen and only happens because it's a duplicate. This is how the script determines what is a duplicate or false mount point.

2. Open another Terminal window and do sudo rm -rf "/Volumes/Macintosh " (I'm putting this path in quotes to include the 2 spaces)

This will delete the false mount point.

3. Quit and relaunch the script. Things should work as expected, now.

 

On a side note, I notice you have some undeleted trash on one of the RAID helpers, "Macintosh 1"

The log shows: (Trash size: 72.7M)

This isn't a problem now, but eventually, if you continue to delete files without emptying the trash, you'll run out of space on that partition and RAID copies will fail, prompting one to wonder what happened. This happened to me many times. LOL! That's why the script keeps track of the trash and partition sizes. You'll have to View that helper Partition for the trash to show the contents and allow you to empty it. At any rate, the script will inform you if the copies failed and revert to original backup files.

 

In the script Preferences, one can disable the "Always check for duplicate mount points" option. I strong suggest not doing so, unless one strongly feels the false mount point is, err, false. This is the only real warning the script provides that indicates a serious issue affecting file access. All sorts of things break with this issue not being resolved.

 

Now, I'm going to look into why the Duplicate Mount Point warning didn't appear when you had the RAID volume selected as the target. I'll get this fixed so that it'll be even more annoying.

Unfortunately, there really isn't any other way around this mount point issue. When you're mounting/unmounting volumes many times and something goes wrong (crash, unexpected quit in the middle of an operation, etc.), the mount point doesn't get cleared properly and gets stuck there, even after a reboot. It has to be cleared manually. And, there is no way I will implement a routine in the script, as simple as it would be, to automatically delete a false mount point (like rm -rf /Volumes/Hard_Drive). That's asking for too much trouble. Ha!

 

best of wishes!

MAJ

 

Ah MAJ,

Where would I be without all your help. I followed your instructions and got it all taken care of (even cleaning out my trash just forgot to clean up after some remodeling The mutiple mount point warning only comes up when I selected one of my other disks to mount. I thought it was just upset at my helper partitions, hence why I named them differently. Oh well all is working very smoothly now thank you for all the pointers.

[digital_dreamer]

Sorry MAJ, I wiped the drive before reinstalling with /Extra method, as I wasn't sure what the problem was when I got stuck with 10.6.4, and started from scratch again. Next install I'll provide the log. Sorry again.

 

Edit scratch that, seems I copied the hackinstaller folder with the original log in it. It's a bit abridged (since I reinstalled 3 or 4 times), but I think the section that interests you is there.

 

Kind regards,

 

Thorgal

Thorgal,

Is your "Snow Leopard" drive USB?

I see where the issues were when you had the EFI partition setup. Oddly, there's no permissions data on any files on that EFI partition. They are all "unknown," which explains why you had the boot cache issues.

 

I'm still looking further into it and believe it may be related to Volume ownership issues. I'll be doing some more tests and may offer you a beta update to see if the problem persists. It's rather odd and I've never run into it before.

 

MAJ

[Thorgal]

Thorgal,

Is your "Snow Leopard" drive USB?

I see where the issues were when you had the EFI partition setup. Oddly, there's no permissions data on any files on that EFI partition. They are all "unknown," which explains why you had the boot cache issues.

 

I'm still looking further into it and believe it may be related to Volume ownership issues. I'll be doing some more tests and may offer you a beta update to see if the problem persists. It's rather odd and I've never run into it before.

 

MAJ

 

Yes, I'm installing from my Mac on an external USB case containing the (SSD) hard drive. The SSD itself is SATA though. After installing I remove the SSD from the casing and put it inside the PC for booting.

[digital_dreamer]

Yes, I'm installing from my Mac on an external USB case containing the (SSD) hard drive. The SSD itself is SATA though. After installing I remove the SSD from the casing and put it inside the PC for booting.

Odd stuff.

Here's what I'm seeing from the log:

You have a properly partitioned GUID (SSD) drive, with the typical 200MB EFI partition (disk1s1):

/dev/disk1
  #:					   TYPE NAME					SIZE	   IDENTIFIER
  0:	  GUID_partition_scheme						*64.1 GB	disk1
  1:						EFI						 209.7 MB   disk1s1
  2:				  Apple_HFS Snow Leopard			63.8 GB	disk1s2

However, the first time the script attempts to reformat/erase the EFI partition to make it a "HFS+" volume with the following command:

diskutil eraseVolume "HFS+" "EFI" /dev/disk1s1

 

The log spits out this info (note the size of the EFI partition - disk1s1):

Erase and format EFI partition as HFS+
Started erase on disk1s1 NO NAME
Unmounting disk
Erasing
Initialized /dev/rdisk1s1 as a 2 GB HFS Plus volume
Mounting disk
Finished erase on disk1s1 EFI
Volume EFI on disk1s1 unmounted

Notice that the EFI partition is no longer 200MB as per log, but initialized as 2GB!

I thought that the low level format of the command is not creating a aligned data structure, as it's using a different block size.

However, one will note that I'm using the "Disk Utility" CLI to format this partition, the same utility (with GUI) that created the proper GUID drive to begin with. So, the capability is there to do it right, I had assumed -- I just hadn't been able to tap into it, yet.

As a side note, the "Started erase..." message contains "NO NAME" on the line, as if it didn't understand that we wanted the name to be "EFI", as indicated in the command line. This is something that has never appeared before on any other drive types, AFAIK.

 

As it turns out, the erase of the EFI partition didn't occur, as the disk list would show later.

 

Later, a second bootloader install with EFI erase proceeds as normal::

Erase and format EFI partition as HFS+
Started erase on disk1s1
Unmounting disk
Erasing
Initialized /dev/rdisk1s1 as a 200 MB HFS Plus volume
Mounting disk
Finished erase on disk1s1 EFI
Volume EFI on disk1s1 unmounted

However, all data on that EFI partition ends up with "unknown" permissions data, even after it has been explicitly set, as the following shows;

drwxr-xr-x@  7 _unknown  _unknown   238B  2 okt 14:03 Themes/
drwxr-xr-x  12 _unknown  _unknown   408B  2 okt 15:04 _Kexts_For_Extra_Cache/
-rwxr-xr-x   1 _unknown  _unknown   579B  2 okt 14:03 com.apple.boot.plist*
-rwxrwxrwx   1 _unknown  _unknown   1,1K  2 okt 15:04 smbios.plist*

 

I can see that the more the disk is used the more permissions data gets assigned.

It seems there's an inability to accept filesystem metadata reliably. Weird and unpredictable.

 

I'm looking deeper for solutions on the web, as I'm sure we're not the only ones running into this.

As SSDs become more popular, this is a issue that needs a solution.

 

 

Anyhow, thanks for letting ramble incoherently. LOL!

What brand/model is this SSD?

 

MAJ

[Thorgal]

Wow, this seems to be more complex than I imagined.

 

First the brand of the SSD:it's a pretty old spare one I had around : an OCZ Core V2 (64Gb).

 

To explain the reformat of the EFI partition : I was actually trying to install to an older EFI partition made with an earlier script of yours, I think it was the original script even you built for snow leopard (I was an early adopter). The newer script however could not access this old EFI partition saying it wasn't correctly formatted (the old script could still access it without trouble). It was correctly formatted IMO, but I had to start over with a new format in the end.

 

What I'll do : I'll look up the unabridged install log, it contains a second and third install try as well. On every try however I kept running into kept cache trouble every time I was installing to /EFI. There never was a problem to install to /Extra, as mentioned before.

 

Kind regards,

thorgal

[satandole666]

Any reason this wouldn't work on a Gigabyte EX-58 Extreme?

 

I think they are pretty much identical.

[Sidewinder33]

Any reason this wouldn't work on a Gigabyte EX-58 Extreme?

 

I think they are pretty much identical.

 

Someone can certainly correct me if I'm mistaken but I've heard the only difference between the two is the on board cooling.